MAJOR ARTICLE

An International Multicenter Performance

Analysis of Cytomegalovirus Load Tests
Hans H. Hirsch,1 Irmeli Lautenschlager,2 Benjamin A. Pinsky,3 Laura Cardeoso,4 Shagufta Aslam,5 Bryan Cobb,5
Regis A. Vilchez,5,a and Alexandra Valsamakis6
1

Transplantation and Clinical Virology, Department Biomedicine (Haus Petersplatz), University of Basel, Switzerland; 2Department of Virology, Helsinki
University Hospital and University of Helsinki, Finland; 3Department of Pathology, Stanford University, Palo Alto, California; 4Department of
Microbiology, Hospital Universitario de la Princesa, Madrid, Spain; 5Roche Molecular Systems, Inc, Pleasanton, California; and 6Department of
Pathology, The John Hopkins Hospital, Baltimore, Maryland

(See the Editorial Commentary by Caliendo on pages 3745.)

Keywords.

Downloaded from http://cid.oxfordjournals.org/ by guest on March 2, 2014

Background. Quantication of cytomegalovirus (CMV) load is central to the management of CMV infections in
immunocompromised patients, but quantitative results currently differ signicantly across methods and laboratories.
Methods. The COBAS AmpliPrep/COBAS TaqMan CMV Test (CAP/CTM CMV test), developed using the rst
World Health Organization CMV standard in the calibration process, was compared to local assays used by 5 laboratories at transplant centers in the United States and Europe. Blinded plasma panels (n = 90) spiked with 2.186.7
log10 copies/mL and clinical plasma samples from immunocompromised patients (n = 660) were tested.
Results. Observed mean panel member concentrations by site and 95% condence intervals (CIs) of the data
combined across sites were narrower for CAP/CTM CMV test compared with local assays. The 95% CI in log10
copies/mL of the combined data per panel member for CAP/CTM CMV test vs comparator assays was .17 vs 1.5 at
2.18 log10 copies/mL; .14 vs .52 at 2.74 log10 copies/mL; .16 vs .6 at 3.3 log10 copies/mL; .2 vs 1.11 at 4.3 log10 copies/
mL; .21 vs 1.13 at 4.7 log10 copies/mL; and .18 vs 1.4 at 6.7 log10 copies/mL. In clinical specimens, constant and
variable quantication differences between the CAP/CTM CMV test and comparator assays were observed.
Conclusions. High interlaboratory agreement and precision of CAP/CTM CMV test results across 5 different
laboratories over 4 orders of magnitude suggest that this assay could be valuable in prospective studies identifying
clinical viral load thresholds for CMV treatment.
CMV; viral load; PCR; transplantation; standardization.

Human cytomegalovirus (CMV) infection causes significant morbidity and mortality in the posttransplant
period of both solid organ transplant (SOT) and hematopoietic stem cell transplant (HSCT) recipients [15].

Received 1 June 2012; accepted 4 September 2012; electronically published 24

October 2012.
a
Present afliation: Antiviral Global Project Team, Global Pharmaceutical R&D,
Abbott Laboratories, Abbott Park, Illinois.
Correspondence: Hans H. Hirsch, MD, MS, University of Basel, Department Biomedicine, Haus Petersplatz, Transplantation and Clinical Virology, and Division of
Infection Diagnostics (Institute for Medical Microbiology), Petersplatz 10 CH-4003
Basel, Switzerland (hans.hirsch@unibas.ch).
Clinical Infectious Diseases 2013;56(3):36773
The Author 2012. Published by Oxford University Press on behalf of the Infectious
Diseases Society of America. This is an Open Access article distributed under the
terms of the Creative Commons Attribution License (http://creativecommons.org/
licenses/by/3.0/), which permits unrestricted reuse, distribution, and reproduction in
any medium, provided the original work is properly cited.
DOI: 10.1093/cid/cis900

Seronegative recipients of transplants from seropositive

donors are infected by CMV transmitted through the
transplanted organ or inadvertently through CMVpositive blood products [15]. In CMV-seropositive recipients, reactivation of latent CMV occurs when
CMV-specic immune control is impaired by immunosuppressive drugs and T-celldepleting therapies [15].
The diagnosis of CMV replication and disease in
SOT and HSCT recipients can be made using different
laboratory methods, including histology, pp65 antigenemia, or CMV DNA by quantitative nucleic acid
testing (QNAT) [15]. Culture methods of body uids
and tissue samples are generally slow and are not
quantitative [4]. The pp65 antigenemia test is rapid (1
day time-to-result) but less sensitive than QNAT and
often difcult to perform on severely neutropenic

CMV Load Standardization

CID 2013:56 (1 February)

367

METHODS
CAP/CTM CMV Test Colinearity to the First WHO CMV
International Standard

The CAP/CTM CMV test (Roche Molecular Systems, Inc

[RMS], Branchburg, New Jersey) analytical performance

368

CID 2013:56 (1 February)

Hirsch et al

characteristics, including traceability to the rst WHO CMV

international standard for nucleic acid amplication techniques (NIBSC 09/162) were described previously [13].
In brief, the CAP/CTM CMV test uses primers and probes
targeting a conserved region of the CMV genome (UL54,
virus encoded DNA polymerase) and has a linear quantication range from 150 to 10 000 000 (2.187.0 log10) copies/mL
representing 137 and 9 100 000 (2.146.96 log10) IU/mL, respectively (1 copy = 0.91 IU). Several standards and control
specimens were used during the development of the test to
achieve traceability to the rst WHO CMV international standard as recommended by the Clinical and Laboratory Standards Institute guidelines [14]. The standards included the
WHO CMV standard, RMS CMV secondary standard, RMS
CMV secondary standard source material (CMV strain
AD169), and RMS CMV calibration panel (Lambda CMA1.2).
The standards, the calibration panel, and an independent
CMV clinical specimen were tested at similar levels to determine whether colinearity to the WHO CMV standard
was achieved. Assessment of colinearity was performed to
demonstrate that the WHO standard was commutable
at any given titer throughout the measuring range and to
thereby ensure traceability. The concentration range tested
for the WHO CMV standard was from 500 IU/mL to 50 000
IU/mL (2.705.70 log10 IU/mL), the RMS CMV secondary
standard source material was tested from 500 IU/mL to
10 million IU/mL (2.707.00 log10 IU/mL), the RMS CMV calibration panel was tested from 523 to 9.3 million IU/mL (2.72
6.97 log10 IU/mL), and the independent CMV clinical specimen
was tested from 500 IU/mL to 22 686 IU/mL (2.704.36 log10
IU/mL). The standard and control specimens were demonstrated to be colinearly distributed to the WHO material
across the linear range of the CAP/CTM CMV test (Supplementary Figure 1).

CMV DNA Quantication Tests

CMV DNA quantication with the CAP/CTM CMV test was

compared to in-house tests of record at each of 5 academic
centers including The John Hopkins Hospital (site 2, real-time
PCR based on Artus reagents [Qiagen, Germantown, Maryland]), Hospital Universitario de la Princesa (site 3, Afgene
real-time PCR test [Cepheid, Sunnyvale, California]) University of Basel (site 4, user-dened real-time PCR [UL111a gene
target]), Stanford University (site 5, COBAS AMPLICOR
MONITOR CMV test [RMS]), and Helsinki University Hospital (site 6, user-dened real-time PCR [ pp65 gene target]).
The analytical performance characteristics for commercial and
laboratory-developed quantitative PCR assays have been described previously [1519].

Downloaded from http://cid.oxfordjournals.org/ by guest on March 2, 2014

patients [17]. CMV QNAT, based most commonly on polymerase chain reaction (PCR), has largely replaced conventional methods owing to better overall performance, and clinical
guidelines now recommend the use of these assays for CMV
load monitoring in SOT and HSCT recipients to prevent or to
manage CMV replication and disease [4, 5, 7, 8].
One central issue that has emerged with the use of different
CMV PCR tests is the signicant interassay quantication variability, as demonstrated in multicenter studies with standardized panels [9, 10]. This lack of assay agreement complicates the
management of individual patients who may have testing performed in different laboratories and it has hampered the establishment of broadly applicable quantitative cutoff values that
can be used in clinical decision making, potentially negatively
impacting the management and long-term outcome of patients
at risk of the direct and indirect effects of CMV replication [1,
4, 5]. Therefore, in the clinical management of CMV after
transplantation, there is a signicant unmet medical need for
the development of standardized nucleic acid tests that deliver
comparable quantitative data across different laboratories.
The rst international standard for CMV QNAT has recently
been established by the World Health Organization (WHO)
Expert Committee on Biological Standardization [11]. This
CMV standard should help to improve interassay agreement.
However, assay-specic variability is still expected owing to underlying differences in test constituents, including varying
nucleic acid extraction methods, target-specic amplication efciencies, assay biochemistries, and operator-dependent variability. As recently suggested [12], these residual quantication
disparities could be solved through the widespread availability
of commercial PCR tests that encompass all assay steps (nucleic
acid preparation, reaction setup, calibration, amplication, and
detection) and demonstrate reliable interlaboratory quantication as dened by agreement and precision. Here, we report the
results of a multicenter international study designed to determine the comparability of quantitative data and precision of a
new, fully automated, Food and Drug Administrationapproved
CMV QNAT (COBAS AmpliPrep/COBAS TaqMan CMV Test
[CAP/CTM CMV test]) using a blinded panel across diverse
laboratories and to compare these 2 parameters among the 5
different assays currently used in these laboratories. Agreement
between the CAP/CTM CMV test and the diverse in-house
quantication assays was also dened using clinical plasma
specimens from immunocompromised individuals.

Comparability and Reproducibility of the CAP/CTM CMV Test

and 5 Quantitative PCR Assays

Quantitative Agreement Between CAP/CTM CMV Test and 5

Quantitative PCR Assays Using Clinical Specimens From
Immunocompromised Patients

Agreement between CAP/CTM CMV test and the 5 quantitative PCR assays described above was investigated with plasma
samples collected at the study sites from immunocompromised patients monitored for CMV replication and disease.
Specimens were assayed only at the site that performed original testing. In addition, 403 samples from 135 HSCT recipients participating in the maribavir prophylaxis for prevention
of CMV phase 3 trial (NCT00411645) were provided by ViroPharma, Inc, to the study sites for PCR testing [21]. Only patients with plasma samples with a volume >600 L were
included in this analysis, and due to volume requirements for
in-house testing, site 2 did not quantify samples from the
maribavir prophylaxis trial. Plasma samples from the maribavir prevention trial were randomly distributed to the other 4
study sites. Institutional review board approval was obtained
at each institution for this study.
Statistical Analysis

The precision of log10-transformed valid test results within the

linear range of each assay was estimated at each expected log10
CMV DNA concentration. The log-normal mean and lognormal coefcient of variation (%) and 95% condence intervals (CIs) (including the lower and upper condence limits) for

Figure 1. Comparability of quantitative data across laboratory sites. A

dilution series of cytomegalovirus (CMV) AD-169 was prepared using
cytomegalovirus-seronegative plasma. Geometric means of tested replicates are plotted. Numerical ranges indicate 95% condence intervals of
the means of each panel member at each site. A, COBAS AmpliPrep/
COBAS TaqMan CMV data. Dashed line indicates assay lower limit of
quantication. B, Data from 5 comparator polymerase chain reaction
assays. Abbreviations: CAP/CTM CMV, COBAS AmpliPrep/COBAS
TaqMan CMV Test; CMV, cytomegalovirus.

total variance were calculated using the linear mixed effect

model with site and day/run, and within-run as random effects.
Deming regression analysis of the viral load results for each
local assay vs the CAP/CTM CMV test was performed to evaluate the correlation between the assays overall and by study
site. All the statistical analyses were performed using the statistical software SAS version 9.2.

CMV Load Standardization

CID 2013:56 (1 February)

369

Downloaded from http://cid.oxfordjournals.org/ by guest on March 2, 2014

Comparability and reproducibility of the CAP/CTM CMV test

was studied in comparison with 3 assays based on commercial
reagents, and 2 tests that use laboratory-developed primers
and probes using a panel prepared from a well-characterized
CMV cultured virus stock (strain AD-169, titer assigned by
the COBAS AMPLICOR CMV MONITOR Test). The panel
consisted of 6 dilutions; 150, 550, 2000, 20 000, 50 000, and
5 000 000 copies/mL (2.18, 2.74, 3.3, 4.3, 4.7, and 6.7 log10
copies/mL, respectively). These dilutions covered the dynamic
range of the CAP/CTM CMV test and represent relevant clinical viral load thresholds [17, 13]. The prepared virus stock
dilutions were further diluted in CMV-negative human ethylenediaminetetraacetic acid plasma. Each site tested 15 replicates
of each panel member with CAP/CTM CMV and in-house
tests except site 2 (15 replicates by CAP/CTM CMV/12 replicates of in-house test). The CMV DNA panel was prepared at
RMS and shipped to the study sites labeled with coded sample
identication numbers to ensure that the site study staff was
blinded to the CMV DNA concentration of each panel
member. These experiments were designed in accordance with
guidelines for establishing analytical performance characteristics of QNATs [20].

RESULTS
Comparability and Reproducibility of Quantitative PCR Assays
With a Standardized CMV DNA Panel

Quantitative Agreement Between CAP/CTM CMV Test and 5

Quantitative PCR Assays Using Clinical Specimens From
Immunocompromised Patients

Trends in quantitative disagreement between the CAP/CTM

CMV test and study sites PCR assays were dened by comparing viral loads obtained by the 2 tests on individual plasma
samples from immuncompromised patients. HSCT recipients
comprised 67% (267/396) of the patients studied; these individuals contributed 80% and 76% of the total number of
samples and valid PCR test results, respectively (Table 1). Two
patterns of disparity were observed (Figure 3). CAP/CTM

370

CID 2013:56 (1 February)

Hirsch et al

Figure 2. Reproducibility of quantitative data across laboratory sites.

Standard deviations of replicates from cytomegalovirus (CMV) AD-169
dilution series are plotted. A, COBAS AmpliPrep/COBAS TaqMan CMV
Test precision by study site. Quantiable replicates at 2.18log10 copies/
mL: site 2, n = 2; site 3, n = 0; site 4, n = 6; site 5, n = 1; site 6, n = 4.
Standard deviations at 2.18log10 copies/mL were not calculated for sites
3 and 5. At 2.74log10 copies/mL, all replicates were quantiable at all
sites except site 6 (11/15 quantiable). B, Precision for the 5 comparator
polymerase chain reaction assays. Replicates of each panel member
tested at site 2, n = 12; at sites 36, n = 15. Quantiable replicates at
2.18log10 copies/mL: site 2, n = 2; site 3, n = 15; site 4, n = 11; site 5,
n = 0; site 6, n = 10. Standard deviation at 2.18log10 copies/mL was not
calculated for site 3. At 2.74log10 copies/mL, all replicates were quantiable at all sites except site 4 (0/15 quantiable) and site 5 (9/15 quantiable). Abbreviation: CAP/CTM CMV, COBAS AmpliPrep/COBAS TaqMan
CMV Test.

CMV test yielded lower values than 3 in-house tests (sites 2, 4,

and 6) throughout the measuring range. Bland-Altman analysis demonstrated this constant bias (least squares regression
slope absolute value <.1, with nonsignicant P value). Additionally, for in-house assays at sites 3 and 5, the difference in

Downloaded from http://cid.oxfordjournals.org/ by guest on March 2, 2014

To determine comparability of quantitative data obtained with

CAP/CTM CMV test across laboratories vs participating
centers PCR assays, plasma panels spiked with CMV strain
AD-169 from 2.186.7 log10 copies/mL were tested at the 5
study sites. The level of quantitative agreement was high for
the CAP/CTM CMV test across the different laboratories as
demonstrated by a smaller range of mean concentrations of
panel members by site and narrower CIs of the combined data
per panel member compared to in-house PCR test in clinical
use at the study sites (Figure 1). For CAP/CTM CMV, the
greatest quantitative variability was observed for the lowest
concentration panel member (2.18 log10 copies/mL, the tests
lower limit of quantication); all replicates were detected, but
62/75 (83%) could not be quantied. This panel member was
also variably detected by each comparator PCR assay. The
AMPLICOR assay (site 5) failed to detect 11/15 replicates
(73%), whereas the Afgene CMV trender test (site 3) was
able to quantify 15/15 replicates. Overall, of 72 valid comparator PCR assay results, 12 replicates were not detected, 22 were
below the lower limit of quantication, and 38 (53%) were
quantied (Supplementary Table 1).
Data from experiments with the spiked plasma panel were
also used to compare the precision of the CAP/CTM CMV
test to the in-house PCR assays used by the study sites. The
standard deviation was <0.2 log10 copies/mL for most panel
members across the different sites performing the CAP/CTM
CMV test, with 2 exceptions (2.18 and 6.7 log10 copies/mL for
site 2 and site 6 tests, respectively, Figure 2). Similar reproducibility was observed for the AMPLICOR test (site 5), CMV
real-time PCR (UL111a PCR, site 4) and real-time PCR ( pp65
PCR, site 6) in-house assays. The Artus CMV PCR (site 2)
and Afgene CMV (site 3) trender tests demonstrated greater
imprecision (standard deviations >0.2 log10 copies/mL for
multiple panel members). Coefcients of variation demonstrated similar trends (Supplementary Figure 2).

Table 1. Patient Populations and Number of Samples Used in

the Reproducibility Comparison of COBAS AmpliPrep/COBAS
TaqMan CMV Test and 5 Quantitative Polymerase Chain Reaction
Assays

Patient Population
Solid organ transplant recipients
(n = 107)
Hematopoietic stem cell transplant
recipients (n = 267)
HIV-infected patients (n = 9)
Other immunocompromised patients
(n = 13)b

Total No. of
Samples
Tested

No. of Valid
Test Resultsa

107

71

531

286

9
13

9
12

Abbreviation: HIV, human immunodeficiency virus.

a

Only samples with paired results within the linear range of the polymerase
chain reaction assays were included in the comparison analysis.

Other immunocompromised patients included subjects diagnosed with

hematologic malignancies (n = 9) or autoimmune diseases (n = 4) receiving
immunosuppressive therapy.

Downloaded from http://cid.oxfordjournals.org/ by guest on March 2, 2014

quantication compared to the CAP/CTM CMV test varied

throughout the measuring range. Bland-Altman analysis
further demonstrated this proportional bias (least squares regression slope absolute value >.1, with signicant P value,
Figure 3).
DISCUSSION
The data from this multicenter international study demonstrate that the CAP/CTM CMV test performs consistently
across laboratories, from the perspective of quantitative agreement and reproducibility. Furthermore, the nding of constant
and variable quantication differences among PCR assays currently used at the participating centers compared with the
CAP/CTM CMV test underscores the challenges in achieving
a general quantitative standardization. Constant bias between
different assays likely reects the use of different calibrators in
assays that are otherwise functionally similar. In these instances, interassay agreement may improve with adoption of a calibrator based on the international standard; alternatively, a
conversion factor can be applied to normalize data if the use
of a different calibrator is not feasible. These fairly easy adjustments are unlikely to improve agreement between assays
whose functionality is sufciently different to result in variable
quantication differences throughout the measuring range.
For these assays (with variable quantication), assay traceability to the international standard alone will not adequately
correct for these types of differences. Instead, these assays
must also demonstrate colinearity to the international

Figure 3. Agreement in cytomegalovirus (CMV) DNA load measurement

between the COBAS AmpliPrep/COBAS TaqMan CMV Test (CAP/CTM
CMV) and 5 comparator polymerase chain reaction (PCR) assays in plasma
samples from immunocompromised individuals. Left-hand panels, agreement plots with Deming regression lines; dashed line indicates 100%
agreement level. Right-hand panels, Bland-Altman plots (difference in
quantication between the CAP/CTM CMV test and comparator PCR
assays vs mean of the 2 measurements). Solid line, least squares regression; dashed lines, mean differences of +0.5/0/0.5 log10 copies/mL. Abbreviation: CAP/CTM CMV, COBAS AmpliPrep/COBAS TaqMan CMV Test.

standard throughout the assay measuring range. Ideally, this

approach (to calibrate and establish colinearity to the reference
material, eg, the rst WHO CMV international standard)

CMV Load Standardization

CID 2013:56 (1 February)

371

372

CID 2013:56 (1 February)

Hirsch et al

discrete, globally applicable, quantitative predictors of active

disease, and other cutoffs that can be used to determine
relapse risk and adequate treatment duration [12]. Currently,
the burden of dening these cutoffs is placed on individual
laboratories, and as a result, clinically relevant values still vary
from center to center.
The implementation of an international standard and the
availability of commercial QNATs with broad interlaboratory
agreement that are traceable and colinear to the rst WHO
CMV international standard represent a much-needed advancement. As demonstrated here for CAP/CTM CMV,
precise, accurate, and standardized results should allow the
design of multicenter studies to delineate testing algorithms,
including quantitative cutoffs and testing frequencies that
enhance clinical outcomes of CMV infections in HSCT and
SOT patients. In turn, these data can be used as the basis for
management guidelines that should signicantly clarify decision making for clinicians and improve infection outcomes in
at-risk patients.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online
(http://www.oxfordjournals.org/our_journals/cid/). Supplementary materials consist of data provided by the author that are published to benet the
reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or
messages regarding errors should be addressed to the author.

Notes
Acknowledgments. The authors acknowledge Dr Stephen Villano at
ViroPharma, Inc, for providing samples from patients in the marabivir
phase 3 trial (NCT00411645); Michael Forman at Johns Hopkins Hospital;
Dr Alexis Dumoulin and the technicians at the Abteilung Infektionsdiagnostik, University of Basel; Dr Laura Mannonen, Dr Raisa Loginov, Dr
Ilkka Helanter, and the technicians at the Helsinki University Hospital;
Elisea Lomas at the Hospital Universitario de la Princesa; and Dr Tri Do
and Ula Cowen at Roche Molecular Systems, Inc, for assistance provided
in the preparation of the study.
Financial support. This work was supported by Roche Molecular
Systems, Inc, Pleasanton, California.
Potential conicts of interest. H. H. H. has received speaker honoraria
from Roche Diagnostics; S. A. and B. C. are employees of Roche Molecular
Systems, Inc; R. A. V. was an employee of Roche Molecular Systems, Inc;
A. V. has been a member of the Roche Diagnostics Scientic Advisory
Board, has been a consultant for Qiagen, and has received clinical trial
funding from Roche Molecular Systems and Qiagen. All other authors
report no potential conicts.
All authors have submitted the ICMJE Form for Disclosure of Potential
Conicts of Interest. Conicts that the editors consider relevant to the
content of the manuscript have been disclosed.

References
1. Gerna G, Lilleri D, Furione M, Baldani F. Management of human cytomegalovirus infection in transplantation: validation of virological
cut-offs for preemptive therapy and immunological cut-offs for production. New Microbiol 2011; 34:22954.

Downloaded from http://cid.oxfordjournals.org/ by guest on March 2, 2014

should be used whenever samples are evaluated in patients

who are monitored for CMV DNA as part of the management
of SOT and HSCT recipients.
Monitoring CMV DNA has become critical for the early
identication of viral replication for preventing progression to
disease in the posttransplant period, and for monitoring the
response to antiviral treatment in patients with CMV replication and disease. In SOT, universal prophylaxis and preemptive treatment approaches are both used to prevent CMV
disease [1, 2, 4, 5]. Although both approaches are currently
viewed as equivalent, the advantages of prophylaxis are prevention of replication and disease in the immediate posttransplant period and elimination of the need for viral load
monitoring during prophylaxis, particularly for the high-risk,
CMV-seropositive donor/CMV-seronegative recipient SOT.
However adverse events associated with prolonged antiviral
drug administration have limited the utility of this approach
in some patients. Also, viral load monitoring may still be
useful in some patients in whom drug-resistant viruses are
suspected to emerge when treated with lower antiviral doses
in an attempt to mitigate drug side effects. Finally, the onset
of late CMV disease has been observed in up to 29% of SOT
recipients after prophylaxis cessation [1, 2, 4].
Viral load monitoring is the key feature of preemptive
therapy. In this alternate approach, antiviral therapy is initiated before the onset of CMV disease when viral load measurements reach a predictive threshold. Advantages of preemptive
therapy include a smaller proportion of treated patients, shortened therapeutic duration, lowered costs associated with posttransplant medications, and reduced occurrence of drug
toxicity ( primarily bone marrow suppression). Avoidance of
marrow suppression is the major rationale for the use of preemptive strategies in allogeneic HSCT in the preengraftment
period. Disadvantages of the preemptive antiviral strategy
include risk of disease prior to treatment initiation in individuals with fast replication, indirect effects of CMV replication
(in the absence of disease) on allograft survival and mortality,
and laboratory costs due to more frequent viral load monitoring. Perhaps one of the most signicant drawbacks to this approach is the lack of generally established quantitative cutoffs
that are predictive of CMV disease in recipients of different
allografts due in large part to the lack of standardized quantication assays that perform comparably across laboratories.
The lack of standardized CMV load assays has also complicated CMV management in other ways. For individual patients, quantication disparities across laboratories dictate that
a single laboratory should be used for viral load testing, so
that results can be accurately interpreted. In addition, although higher viral loads have been shown to correlate with an
increased risk of disease in SOT and HSCT patients [1, 4, 5],
the lack of standardization has hampered the development of

12. Kraft CS, Armstrong WS, Caliendo AM. Interpreting quantitative cytomegalovirus DNA testing: understanding the laboratory perspective.
Clin Infect Dis 2012; 54:17937.
13. Roche Molecular Systems, Inc. COBAS TaqMan CMV Test FDA approved package insert. Branchburg, NJ: Roche, July 2012.
14. Clinical and Laboratory Standards Institute (CLSI). Metrological traceability and its implementation; a report. CLSI document X5-R, Vol
26, No 10. ISBN 1-56238-598-4.
15. Abbate I, Finnstrom N, Zaniratti S, et al. Evaluation of automated extraction system in combination with Afgene CMV tender for CMV
DNA quantitative determination: comparison with nested PCR and
pp65 antigen test. J Virol Meth 2008; 151:615.
16. Koidl C, Bozic M, Marth E, Kessler HH. Detection of CMV DNA: is
EDTA whole blood superior to EDTA plasma? J Virol Meth 2008;
154:21012.
17. Sia IG, Wilson JA, Espy MJ, Paya CV, Smith TF. Evaluation of the
COBAS Amplicor CMV monitor test for detection of viral DNA in
specimens taken from patients after liver transplantation. J Clin Microbiol 2000; 38:6006.
18. Piiparinen H, Hockerstedt K, Gronhagen-Riska C, Lautenschlager I.
Comparison of two quantitative CMV PCR tests, COBAS Amplicor
CMV Monitor and TaqMan assay, and pp65-antigenemia assay in the
determination of viral loads from peripheral blood of organ transplant
patients. J Clin Virol 2004; 30:25866.
19. Beckman C, Dumoulin A, Rinaldo CH, Hirsch HH. Comparison of a
UL111a real-time PCR and pp65 antigenemia for the detection of cytomegalovirus. J Med Virol 2011; 83:214350.
20. Clinical Laboratory and Standards Institute (CLSI). EP9-A2. Method
comparison and bias estimation using patient samples; approved
guideline. 2nd ed. 2002.
21. Marty FM, Ljungman P, Papanicolaou GA, et al. Maribavir prophylaxis for prevention of cytomegalovirus disease in recipients of allogenic
stem cell transplants: a phase 3, double-blind, placebo-controlled, randomized trial. Lancet Infect Dis 2011; 11:28492.

CMV Load Standardization

CID 2013:56 (1 February)

373

Downloaded from http://cid.oxfordjournals.org/ by guest on March 2, 2014

2. Humar A, Snydman D. Cytomegalovirus in solid organ transplant

recipients. Am J Transpl 2009; 9(suppl 4):S7886.
3. Fedorko DP, Yan SS. Recent advances in laboratory diagnosis of human
cytomegalovirus infection. Clin Immunolog Revs 2002; 2:15567.
4. Kotton CN, Kumar D, Caliendo AM, et al., Transplantation Society
International CMV Consensus Group. International consensus guidelines on the management of cytomegalovirus in solid organ transplantation. Transplantation 2010; 7:77995.
5. Boeckh M, Ljungman P. How I treat cytomegalovirus in hematopoietic cell transplant recipients. Blood 2009; 113:57119.
6. Halfon P, Berger P, Khiri H, et al. Algorithm based on CMV kinetics
DNA viral load for preemptive therapy initiation after hematopoietic
cell transplantation. J Med Virol 2011; 83:4905.
7. Preiksaitis JK, Brennan DC, Fishman J, Allen U. Canadian Society of
Transplantation Consensus Workshop on Cytomegalovirus Management in Solid Organ Transplantation nal report. Am J Transplant
2005; 5:21827.
8. Andrews PA, Emery VC, Newstead C. Summary of British transplantation society guidelines for the prevention and management of CMV
disease after solid organ transplantation. Transplantation 2011;
92:11817.
9. Pang XL, Fox JD, Fenton JM, Miller GG, Caliendo AM, Preiksaitis JK,
for the American Society of Transplantation Infectious Diseases Community of Practice and the Canadian Society of Transplantation. Interlaboratory comparison of cytomegalovirus viral load assays. Am J
Transplant 2009; 9:25868.
10. Caliendo A, Shahbazian MD, Schaper C, et al. A commutable cytomegalovirus calibrator is required to improve the agreement of viral
load values between laboratories. Clin Chem 2009; 55:170110.
11. Freyer JF, Heath AB, Anderson R, Minor PD, and the collaborative
study group. Collaborative study to evaluate the proposed 1st WHO
international standard for human cytomegalovirus for nucleic acid
amplication-based assays. WHO/BS/10.2138, pp 140. Geneva, Switzerland, 1822 October 2010.